After having practiced the CAM art and having tuned the machine on wood, I’m finally trying to cut carbon fiber (3k, twill weave, 2mm thick sheet) but cannot get a decent job done. I’m using a 2 mm ball nose, two-flute upcut bit, 1/8" (3.175 mm) shank. My spindle maxes out at ~12.000 RPM but the last 1k makes it whine and shake a little, so I usually run it at 10.000 RPM tops.
I’m cutting 3mm deep to compensate for the ball nose bit - that’s the only 2mm one I have at the moment, I’ll order some PCB/fiberglass milling (diamond pattern) bits soon.
Here are some pictures of my workpiece. The pieces underlined in red belong to the first cut, the ones in yellow to the second; videos of the first cut are included.
Feed
Plunge
Speed
DOC
1st cut
5 mm/s
1 mm/s
9000 RPM
1 mm
2nd cut
2.5 mm/s
0.5 mm/s
10000 RPM
0.5 mm
The first cut broke the bit right at the end of the job, so I thought that I was trying to go too fast. Slowing down made the second cut a little cleaner, but there still are signs of flex/slop and this time the bit did not even make it through the CF. After the cut the bit was cool and the spindle was not overheated.
The material is rather stiff, but it is plausible that it flexes down when the bit plunges and hence the problems with the cut’s depth, especially when restarting after an holding tab - should I come up with a better fastening solution, maybe a support in the center?
Can anybody point me in the right direction to start getting half-decent results?
Carbon fiber is not any different than cutting regular plastic other than It is extremely dangerous to breathe the dust (think asbestos but worse) , and it wears out bits really fast.
I suspect all your issues are with trying to use a ballend that has already went dull. Get a single flute endmill.
Hey Ryan, thanks for the lightning fast response! Sorry for the double post, trying to embed the Imgur album took multiple tries. Feel free to delete either one of the two threads.
I’ll order a new bit straight away and report back. I wanted to try cutting using what I had, but being impatient rarely brings good results
I’m cutting the carbon underwater (the album I linked above includes two short videos), so after the cut I “just” have to safely dispose of that liquid and thoroughly clean the parts from any dust residues. So far it has proven a good method, after the first cut I’ve also 3D-printed a shroud for the spindle to somehow contain spills.
I’ve attached the gcode for the second cut, I don’t have the one for the first anymore but can reproduce it if needed. The clamps hold the sheet firmly in the XY plane, but I did not pay much attention to flex in the Z direction when designing them. How would you suggest to hold the sheet? Maybe double sided tape applied before adding water?
Shouldn’t The resin by itself is similar to acrylic, but the Carbon is like grinding rocks. Fiberglass should be the same but teh bit will last a little longer.
Looking at the Gcode, it seems good. You are slotting with a ball end, try punching all the way through and past by a few mm in a single pass. The cuts are pretty wonky though. Maybe the plate is just completely flexing, do you have something under the middle or are you just holding the corners? If not use some wood or HD foam.
There’s nothing under the middle, I trusted the sheet not to flex much. Or at least I trusted it when trying to bend it by hand… I’ll make a shim and try another cut with the ball end but with a single pass. Which feed & plunge would you recommend for “one-passing” this 2mm sheet?
I’ve made a spacer to support the center of the sheet, now I’m waiting for some proper bits to retry the cut.
Maybe I’ll make a taller enclosure for the sheet with a wooden spoilboard inside it, but I’m not too keen on letting MDF or pine (those are what I have at hand at the moment) suck in all of that sweet carbon fiber dust.
I could 3D-print bigger corner blocks, screw the CF sheet in with a washer and use the PLA as a spoilboard… That does not sound like a good solution for the long term though, PLA could melt on the bit.
Things improved a lot with the new bits - there’s a reason they make them, after all… Here are some pictures of the cut plus a bonus shot of the shroud I made to contain spills. It works quite well, I was not expecting it!
There’s still some slop in the cut, I’ve carefully inspected the corner mounts and they indeed developed ~1mm of slack. I think I’ll get some acrylic and cut a pattern into it to hold a few nuts and bolts to keep the sheet steady. @Barry99705 suggested to use insulation foam, but the one I have at hand is too thick and I don’t think it wouldn’t absorb water.
Once you get that material mounted solidly you should be able to cut much faster but you absolutely need something under it and it can not move at all or you will break a bit.
You are close but need to find some HD foam or something.
Any details on the tray you’re using? I’ve been wanting to make something similar. For workholding I was thinking about double sided tape to cheap HPDE cutting boards as a waterproof spoilboard you could bolt down to your water tray somehow.
I’ve found a 5mm black PVC sheet around the house, I’ll use that as a spoilboard inside the tray since the speeds should be adequate to avoid getting melt plastic on the bit for the last pass. I’ll lower my extra cutting depth, maybe only using 0.1 or 0.2mm more than the CF sheet’s thickness.
It is a very simple domestic plastic tray; I cut it to size and taped the edges, nothing fancy. It is held down to the bed with double sided tape and so far has kept up with some abuse. If I had to start over, I’d probably get some thick acrylic sheets and make a proper tray/box, sealing the edges with silicone and placing all the needed fastening holes directly in it. Getting a tray this low off the machine is kinda uncomfortable when it’s filled with polluted water that you’re trying not to spill everywhere… Maybe I should try to remove some of it with a syringe, first.
The supports at the corners were 3D-printed, but they don’t work well and need to be redesigned if I want to keep using that system. There’s also another printed rectangular support under the middle of the sheet.
The new tray base should be much better, I think I’ve nailed it with the nut slots in the PVC - this way it will adapt to the specs of the CF sheets that I’m using since they have a ~3mm tolerance on the reference dimensions. Using the machine to make its own accessories/mounts really reminds me of my first spools of filament used in my 3D printer
In the next few days I think I’ll try another cut, I suppose the feed should be adequate to avoid damaging the underlying PVC to a noticeable extent. I’m thinking of plunging an extra 0.25mm into the PVC spoilboard, is that too much? Now that the CF is secured and can’t flex I could maybe get away with 0.1mm.
Today I tried cutting several pieces and the overall result was quite nice. I now have a problem, though, with the helically drilled holes: the spindle’s mount/assembly/everything flexes way too much when cutting such holes, and what should be a 3mm hole comes out as a wonky 2mm one (since I’m using a 2mm bit).
The speeds used were the ones above, I’ll try shallower passes of 0.5mm. Are there any machining tips I’m not aware of to bring the cut time down a little? I’ll need to cut a few of these pieces, and so far they’ve been pretty tedious to machine - 4x of them took ~50mins with the old feeds & speeds.
Finishing passes are a new tool to me, how would I benefit of them? My edges are already quite clean, I never thought of using them for speed gains.
The spindle flexes noticeably only in helical drills, I tried to move it by hand but it seems to be firmly locked in place. Could it be that, since it gets hot - I'll measure its temperature next time - during use, it softens the PLA mounts? I thought the flex was due to too much force being applied on the bit (e.g.: too fast), but I didn't notice this behavior on wood.
Thanks for great support Ryan, the MPCNC is amazing and it is awesome that you help its users out so much
??? can you get that on video, nothing should move that you can see with your eye. You tool should not get hot enough to soften PLA if it does something is wrong and it would fall out. This is a huge problem, we are worried about 0.01mm movement if you can see it moving you are lucky your are not just breaking your bits.
How about a picture of your build. This will answer many many questions I have. From the video the first thing I am seeing is the bottom of your collet should be just a bit lower than the bottom of your Z rails. This does not show much of the build itself though.
